PDF to PNG — Lossless, Pixel-Exact Pages

No. PNG is a lossless format, so the pixels written to the file are exactly the pixels the page was drawn as. Nothing is approximated and nothing is thrown away. The only thing that limits fidelity is the resolution you render at — below a certain point small type stops being legible no matter how faithfully each pixel is stored, because the detail was never drawn in the first place.

No, and it is worth being clear about this because PNG is the format people associate with transparency. A PDF page is defined as ink on white paper rather than as a floating object, so every page is drawn on an opaque white ground before anything else lands on it. What you get is a solid white sheet with the page content on it. If you need a logo or a signature isolated from its background, convert the page here first and then take the result to a background remover.

Because of how the two formats treat sharp edges. Lossy compression works on blocks of pixels and smooths away differences inside each block, which is invisible in a photograph but shows up as a faint gray shadow along the boundary between solid black type and white paper. PNG stores each pixel exactly, so a letter edge stays a clean step from black to white. On a page of small type at a modest resolution that is a visible difference, not a theoretical one.

Never. Reading and rendering both happen inside this browser tab using its own PDF engine, and the file is not transmitted at any point. The tool also works with your network disconnected once the page has loaded, which is the simplest way to prove it to yourself. This matters because the documents people rasterize tend to be contracts, statements and scanned identity papers rather than holiday photographs.

Pick PNG when the page is made of hard-edged marks: text, tables, line drawings, charts, screenshots, engineering diagrams, sheet music, anything with thin rules or fine hatching. Pick JPG when the page is essentially a photograph, which includes most flatbed scans of printed matter, because photographic detail hides lossy artifacts well and the saving is large. A page that mixes a paragraph of text with one photograph is the awkward case, and PNG is the safer default there.

Yes, and this is where lossless output earns its keep. Every time a lossy image is opened, altered and saved again it degrades a little further, so cropping and re-saving a page four times leaves visible damage. A PNG survives that cycle untouched. Slide decks, documentation and anything that will be annotated later are all better served by PNG, which is also why design and word-processing software imports it without complaint.

Yes to both. A scanned page is rasterized like any other, though a scan is already a photograph of paper so PNG will store its sensor noise and paper texture faithfully and honestly rather than smoothing them away. Chinese, Japanese, Korean, Arabic, Hebrew, Devanagari and other scripts render correctly because the character maps and the standard font set are served from this site rather than fetched from elsewhere, so nothing about your document is disclosed in order to draw it.

They stop being interactive, because an image cannot hold behavior. Hyperlinks become underlined words that do nothing, a form field becomes a picture of a box, and text that could be selected and searched becomes part of the picture. Screen readers can no longer read the page either, which is the accessibility cost of any rasterization. Keep the original PDF as the authoritative copy and treat the images as a faithful likeness of it.

Because you have swapped a description for a photograph of that description. A PDF page stores instructions such as draw this glyph at this position, which is compact regardless of how large the page is. A rasterized page instead stores a value for every single pixel, and storing them exactly is the definition of lossless. Expect several times the size of the original page, and treat that as the price of exact fidelity rather than as something to tune away.